Bortezomib in the Rapid Reduction of High Sustained Antibody Titers in Disorders Treated with Therapeutic Protein: Lessons Learned from Pompe Disease

©American College of Medical Genetics and Genomics ORIGINAL RESEARCH ARTICLE

Bortezomib in the rapid reduction of high sustained antibody titers in disorders treated with therapeutic protein: lessons learned from Pompe disease

Suhrad G. Banugaria, MBBS1, Sean N. Prater, MD, MRes1, Judeth K. McGann, MD2, Jonathan D. Feldman, MD2, Jesse A. Tannenbaum, MD2, Carrie Bailey, BS, CCRC3, Renuka Gera, MD4, Robert L. Conway, MD4, David Viskochil, MD, PhD3, Joyce A. Kobori, MD5, Amy S. Rosenberg, MD6 and Priya S. Kishnani, MD1

Purpose: High sustained antibody titers complicate many disor- titers, a regimen based on bortezomib (Velcade) was used in com- ders treated with a therapeutic protein, including those treated with bination with rituximab, methotrexate, and intravenous immuno- enzyme replacement therapy, such as Pompe disease. Although globulin. enzyme replacement therapy with alglucosidase alfa (Myozyme) in Pompe disease has improved the prognosis of this otherwise lethal Results: The treatment regimen was well tolerated, with no obvi- disorder, patients who develop high sustained antibody titers to alg- ous side effects. Patient 1 had a 2,048-fold, and patients 2 and 3 each lucosidase alfa enter a prolonged phase of clinical decline resulting in had a 64-fold, reduction in anti-alglucosidase alfa antibody titer, with death despite continued enzyme replacement therapy. Clinically effec- concomitant sustained clinical improvement. tive immune-tolerance induction strategies have yet to be described Conclusion: The addition of bortezomib to immunomodulatory in the setting of an entrenched immune response characterized by regimens is an effective and safe treatment strategy in infantile Pompe high sustained antibody titers, wherein antibody-producing plasma disease, with potentially broader clinical implications. cells play an especially prominent role. Genet Med 2013:15(2):123–131 Methods: We treated three patients with infantile Pompe disease experiencing marked clinical decline due to high sustained antibody Key Words: antibodies; bortezomib; enzyme replacement therapy; titers. To target the plasma cell source of high sustained antibody glycogen storage disease type II; immune modulation

Background immune response to ERT are an ongoing challenge in Pompe Protein replacement therapies have prolonged the survival and disease. Prior studies have demonstrated the negative impact of improved clinical outcomes of patients suffering from a multi- high sustained anti-rhGAA IgG antibody titers (high sustained tude of disorders. However, therapeutic proteins are potentially antibody titers; HSATs) on clinical outcomes in IPD.2–4 More immunogenic, eliciting antibody responses that reduce efficacy.1–5 recently, there are reports of the negative impact of HSATs in Infantile Pompe disease (IPD) results from the deficiency of lys- adults with late-onset Pompe disease.13,14. osomal acid α-glucosidase (GAA). Classic IPD is characterized Some of the patients treated with alglucosidase alfa also by cardiomyopathy, hypotonia, respiratory insufficiency, and, if develop IgE antibodies and appear to be at a higher risk for untreated, death before 2 years of age.6–8 Atypical, or nonclassic, anaphylaxis and severe allergic reactions.15 IgE is measured in patients with IPD present in infancy and typically do not have the setting of a hypersensitivity reaction and is not routinely severe cardiomyopathy; and in some instances there is no car- tested. The longer term therapeutic efficacy of alglucosidase alfa diac involvement. Untreated patients with atypical IPD also have appears to be more strongly associated with anti-rhGAA IgG rapid disease progression, becoming wheelchair bound and/or antibodies. The impact of IgE antibodies on clinical outcomes, ventilator dependent in the first few years of life.9 The availabil- in the absence of anaphylaxis, is still unclear and requires fur- ity of enzyme replacement therapy (ERT) with Chinese hamster ther investigation. In addition to the antibody response, ERT ovary cell-line-derived recombinant human acid α-glucosidase has also been shown to induce a T-cell response.16 (rhGAA, alglucosidase alfa, Myozyme, Genzyme, Cambridge, IPD serves as an excellent model with which to evaluate MA) has led to significant improvements in overall survival and immune tolerance induction (ITI) protocols as it is a rapidly pro- other clinical outcomes.10–12 However, complications from the gressive disease and any clinical interventions or factors altering

1Division of Medical Genetics, Department of Pediatrics, Duke University Medical Center, Durham, North Carolina, USA; 2Deparment of Pediatrics, Kaiser Permanente, Santa Clara, California, USA; 3Division of Medical Genetics, Department of Pediatrics, University of Utah, Salt Lake City, Utah, USA; 4Department of Pediatrics/Human Development, Michigan State University, East Lansing, Michigan, USA; 5Department of Pediatrics, Kaiser Permanente, San Jose, California, USA; 6Division of Therapeutic Proteins, Office of Biotechnology Products, Center for Drug Evaluation and Research, United States Food and Drug Administration, Bethesda, Maryland, USA. Correspondence: Priya S. Kishnani ([email protected]) Submitted 5 June 2012; accepted 27 July 2012; advance online publication 11 October 2012. doi:10.1038/gim.2012.110

Genetics in medicine | Volume 15 | Number 2 | February 2013 123 ORIGINAL RESEARCH ARTICLE BANUGARIA et al | Bortezomib in Pompe disease efficacy of ERT tend to manifest rapidly via well-defined clini- explained by the hitherto use of agents that do not target anti- cal end points. Recently, emerging clinical evidence regarding body-producing plasma cells.20,24 Bortezomib is a proteasome the effect of the immune response to ERT in IPD has spawned inhibitor that targets both short- and long-lived plasma cells, and investigations into new approaches toward ITI designed both is an agent of potential benefit in eliminating HSAT.25,26 to preclude an immune response in the ERT-naive setting and Herein we present three patients with IPD treated with ERT to eliminate the immune response in patients with IPD who who, despite initial clinical improvement, subsequently declined had only recently commenced ERT.17,18 Despite attempts at ITI following the development of HSAT. All three patients were in patients with IPD who developed HSAT, success has been treated with a bortezomib-based immunomodulatory regimen. elusive.19,20 Combinations of various drugs, such as rituximab, cyclophosphamide, and intravenous immunoglobulin (IVIG), Methods and plasmapheresis in IPD with HSAT and in hemophilia have We included patients who met criteria for IPD (classic or atypi- shown little success in reducing HSAT.19–23 Failure of these agents cal) with anti-rhGAA IgG antibody titers of ≥1:51,200 on two to reduce and sustain a reduced antibody titer can potentially be different occasions at/or beyond 6 months post-ERT initiation.

a Alglucosidase Alglucosidase Alglucosidase alfa – 20 mg/kg every other week alfa – 40 mg/kg every alfa – 30 mg/kg other week every week 250,000 400

360.3 1:204,800 350 )

200,000 2

300 x (g/m y titer 278.6 150,000 250 228.6

1:102,400 200 100,000 193 195

149 150 1/ Anti-rhGAA antibod 1:51,200 1:51,200 135

50,000 Left ventricular mass inde 108 100 1:25,600 85 72 1:6,4001:3,200 1:400 1:100 0 50 01020530 40 060970 80 0 100 110 120 130 140 150 160 Weeks on ERT

Cyclophosphamide (250 mg/m2 IV) Rituximab (375 mg/m2 IV) Bortezomib (1.3 mg/m2 IV) Methotrexate (15 mg/m2 orally every other week) IVIG (400−500 mg/kg IV monthly) Antibodies LVMI

b Alglucosidase alfa – 20 mg/kg every other week 900,000 210 200.9 1:819,200 800,000 190 ) 2 700,000 170

y titer 158 600,000 x (g/m 150 139 500,000 1:409,600 1:409,600 130 400,000 110 300,000 95 105 1:204,800 1:204,800 1:204,800 82.5 90 200,000 77.8 1/Anti-rhGAA IgG antibod

1:102,400 Left ventricular mass inde 1:102,400 70 100,000 1:51,200 1:6,400 60.3 1:12,800 0 50 01020 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Weeks on ERT Rituximab (375 mg/m2 IV) Methotrexate (15 mg/m2 SC)

Bortezomib (1.3 mg/m2 IV) IVIG (400−500 mg/kg IV monthly) Antibodies LVMI

124 Volume 15 | Number 2 | February 2013 | Genetics in medicine Bortezomib in Pompe disease | BANUGARIA et al ORIGINAL RESEARCH ARTICLE

Patients 1 and 2 met the criteria for classic IPD: ≤1% of nor- detector and tandem mass spectrometry (electrospray ioniza- mal GAA activity (in skin fibroblasts and/or muscle biopsy), tion mass spectrometry), as previously described.30. After dis- cardiomyopathy (left-ventricular mass index (LVMI) ≥65 g/m2 cussion with the Northern California Regional Kaiser Genetic by echocardiogram), and presentation within the first year of Diseases Treatment Advisory Board, and after obtaining written life. Patient 3 had atypical IPD without cardiomyopathy and parental informed consent, a trial of a bortezomib-based regi- presented in the first year of life. Cross-reactive immunologic men was commenced in patient 1. The parents of patients 2 and material (CRIM) status was determined as described earlier 3 agreed to physician-directed use of bortezomib as a life-saving based on the reactivity of a pool of monoclonal and polyclonal measure. Bortezomib was available in both hospitals’ formulary anti-GAA antibodies capable of recognizing both native and for pediatric use in refractory leukemia. Thebortezomib-based recombinant GAA.27,28 A patient was designated as CRIM- regimen is outlined in Figure 1a–c for patients 1, 2, and 3, positive if any of the GAA protein forms (unprocessed precur- respectively. sor band at 110 kDa or any of the processed forms) were detectable on western blot analysis; a patient was designated as CRIM Case description and findings negative if none of these protein forms was detectable on west- Patient 1. This patient, a 4-year-old Caucasian male, was diag- ern blots (processed and unprocessed).3 Anti-rhGAA antibody nosed with IPD at the age of 5 months. Patient demographics, titers were serially evaluated by Genzyme, as recommended in CRIM status, and mutation data are shown in Table 1. Disease- the Myozyme package insert.29 The fold reduction in antibody associated signs and symptoms were first noted at age 3 months titer was calculated by dividing the initial titer value by the cur- and included hypotonia as well as feeding and sucking diffi- rent antibody titer value. Urinary glucose tetrasaccharide (Glc4), culties. He had significant cardiomegaly at 5 months. Baseline a biomarker for overall glycogen burden in skeletal muscle, was (i.e., before initiation of ERT) clinical parameters, including determined by high-pressure liquid chromatography with UV cardiac, motor, respiratory, and feeding status, and laboratory

c Alglucosidase alfa – 20 mg/kg every other week 250,000 160

1:204,800 149.2 150 200,000 140 135 130 y titer

150,000 120.4 120

111 110 105.5 (mmol/mol creatinine) 100,000 1:102,400 1:102,400 96.2 100 4 93 90 1:51,200

1/Anti-rhGAP antibod 1:51,200

50,000 80 Urine Glc 72.9 67 1:25,600 1:12,800 66.9 70 1:12,800 1:6,400 1:3,200 0 60 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 Weeks on ERT

Rituximab (375 mg/m2 IV) Methotrexate (15 mg/m2 SC every other week) Bortezomib (1.3 mg/m2 IV) IVIG (400−500 mg/kg IV monthly)

Antibodies Glc4

Figure 1 trends in anti–recombinant human alglucosidase alfa (anti-rhGAA) IgG antibody titers and left-ventricular mass index (LVMI; (a and 2 39 b) or urine Glc4 (c) over time are shown by a solid black line and a dashed gray line, respectively (the upper limit of normal LVMI is 65 g/m ; control range of urine Glc4 for age >3 years is 3 mmol/mol of creatinine). (a) In patient 1, following unsuccessful immunomodulation with two doses of cyclophosphamide (250 mg/m2 i.v.) and four weekly doses of rituximab (375 mg/m2 per dose i.v.) during weeks 86 to 95, monotherapy with bortezomib (Velcade) was initiated. Bortezomib was administered twice weekly (1.3 mg/m2 of body surface area i.v.) according to a standard dosing regimen (days 1, 4, 8, and 11; equivalent to one cycle of bortezomib)29 during weeks 99 and 100 (cycle 1), weeks 110 and 111 (cycle 2), and weeks 127 and 128 (cycle 3) of enzyme replacement therapy (ERT). Monthly intravenous immunoglobulin (IVIG) was administered at the start of the first cycle of bortezomib. Following a second cycle of bortezomib with continued ERT, a total of four weekly doses of i.v. rituximab (375 mg/m2 per dose) were administered in addition to biweekly methotrexate (15 mg/m2) orally. Rituximab was thereafter administered on a monthly basis. (b) In patient 2, following unsuccessful immunomodulation with rituximab (375 mg/m2, i.v.) and methotrexate (15 mg/m2 s.c.), bortezomib was administered as described in patient 1. Following the first cycle of bortezomib, a total of four weekly doses of rituximab (375 mg/m2 per dose, i.v.) were administered in addition to biweekly methotrexate (15 mg/m2, s.c.). Rituximab was thereafter administered on a monthly basis. Monthly IVIG was administered as described for patient 1. (c) In patient 3, bortezomib was administered as described for patient 1. Six such cycles were administered between weeks 88 and 153. Rituximab was administered (375 mg/m2 per dose, i.v.) approximately every month except after the first and third cycle of bortezomib, at which time a total of four weekly doses were administered after the last bortezomib injection. Methotrexate and IVIG were administered as described in patients 1 and 2.

Genetics in medicine | Volume 15 | Number 2 | February 2013 125 ORIGINAL RESEARCH ARTICLE BANUGARIA et al | Bortezomib in Pompe disease

Table 1 Patient demographics, cross-reactive Following the first cycle of IV bortezomib monotherapy, anti- immunologic material status, and mutation data body titers decreased from 1:204,800 to 1:51,200 (Figure 1a). Patient 1 Patient 2 Patient 3 Cycle two of bortezomib was combined with rituximab, meth- Gender Male Male Male otrexate, and IVIG to preclude activation of rhGAA-specific Race Caucasian Caucasian Caucasian naive and memory B- and T-cells and facilitate tolerance induc- CRIM status Positive Negative Positive tion. This treatment resulted in a further decline of the antibody titer to 1:6,400 (Figure 1a). Following a third round of Mutations Allele 1 c.307T>G c.2560C>T c.1655T>C bortezomib in combination with rituximab and methotrexate, c.2481+102 Allele 2 c.1654delC c.1655T>C antibody titers were 1:100 at week 159 of ERT. The decline in _2646+31del antibody titers in this bortezomib-treated patient over time is Age at symptom onset 3 months 2 months <1 month in stark contrast to the persistence of HSATs in CRIM-negative Age at diagnosis 5.0 months 4.0 months 3.5 years and high-titer CRIM-positive patients, as well as in patients Age at ERT start 6.0 months 4.5 months 3.6 years treated with a non-bortezomib-based immunomodulatory reg- 19 CRIM, cross-reactive immunologic material; ERT, enzyme replacement therapy. imen (Supplementary Figure S1 online). The reduction in antibody titers following implementation parameters such as antibody titers and urinary Glc4 levels are of an immunomodulation strategy using bortezomib was in shown in Tables 2–4. At age 6 months, the patient was started close temporal association with a period of marked and pro- on ERT with alglucosidase alfa at 20 mg/kg every other week. longed clinical improvement, including significant improve- Clinical improvement was noted shortly after the initiation ment in cardiac status, decreased ventilatory requirements, and of ERT, including improvements in cardiac and motor status improved motor status (Tables 2–4). The treatment regimen (Tables 2–4). was well tolerated with no apparent side effects. At the time of this report, the patient is no longer receiving bortezomib. Rising antibody titers. From a seronegative status at week 4 of Rituximab dose intervals have been increased to 2 months. He ERT, anti-rhGAA antibody titers continued to rise, peaking at continues on methotrexate and IVIG, the latter as clinically 1:204,800 at week 64 of ERT, and were maintained at 1:204,800 needed. The patient continues to receive alglucosidase alfa infu- through week 86. Concomitant with rising antibody titers was sions at 30 mg/kg every week. a progressive decline in clinical status and increase in urine Glc4 levels (Tables 2–4). The patient became fully ventilator depen- Patient 2. Patient 2 is a 4½-year-old male diagnosed with IPD dent and required tube feeds. Anti-rhGAA IgG antibody titers at 4 months of age. Patient 2 demographics, CRIM status, and and LVMI values over time are shown in Figure 1a. mutation data are shown in Table 1. This patient presented with a cardiac murmur at age 2 months and an echocardiogram Immunomodulation with cyclophosphamide and ritux- showed significant biventricular hypertrophy. Physical exam imab. Due to HSATs and concurrent clinical decline, ITI was at age 3 months showed signs of IPD (Tables 2–4). At age 4.5 initiated. Cyclophosphamide (250 mg/m2 i.v.) monotherapy months, he was started on alglucosidase alfa at 20 mg/kg every was administered at weeks 86 and 92 of ERT, followed by ritux- other week. Clinical improvement was observed after ERT ini- imab (375 mg/m2 i.v.) every week from week 92 to week 95 tiation and included a reduction in LVMI and improved feeding post-ERT initiation (Figure 1), based on published reports.16,17 (Tables 2–4). Despite elimination of a subset of B cells from peripheral blood, as evidenced by a CD19 count of 0%, antibody titers Rising antibody titers. From a seronegative status at baseline remained persistently high at 1:204,800, and the patient con- antibody titers continued to rise and peaked at 1:819,200 at tinued to decline clinically in terms of cardiac, motor, and week 86 of ERT. Concomitant with rising antibody titers was a respiratory status (Tables 2–4). Increasing LVMI resulted in plateau of his gross developmental progress and increased uri- a near-complete left-ventricular cavity obstruction and signifi- nary Glc4 (Tables 2–4). cant left-ventricular outflow tract obstruction. At this time, he Immunomodulation with rituximab and methotrexate. also had an increase in ventilation requirement. Urinary Glc4 levels peaked at 153.3 mmol/mol of creatinine. The patient was Given the rise in antibody titers, plateau in clinical response unable to move his arms or legs and his voluntary motor activ- and his CRIM-negative status, an ITI protocol was initiated at ity was limited to ocular movements. week 87 that included rituximab (375 mg/m2 i.v., every week for 4 weeks and methotrexate (15 mg/m2 s.c.) every other week. Immunomodulation using bortezomib-based regimen This regimen led to an eightfold decrease in antibody titers and subsequent clinical improvement. Given the dire prog- over 29 weeks: 1:819,200 at week 86 to 1:102,400 at week 115. nosis of this child, and the reasoning that rituximab-resis- He maintained adequate physical activity and initially pro- tant, antibody-secreting plasma cells had to be eliminated, gressed in developmental motor milestones. Although there a trial of the bortezomib-based regimen was commenced was an initial phase of improved LVMI, the values continued (Figure 1a). to deteriorate thereafter (Figure 1b, Tables 2–4). Even after his

126 Volume 15 | Number 2 | February 2013 | Genetics in medicine Bortezomib in Pompe disease | BANUGARIA et al ORIGINAL RESEARCH ARTICLE

Table 2 Clinical and laboratory parameters over time for the three patients treated with bortezomib-based regimen (patient 1) Patient 1 Immunomodulation attempt with Initial improvement Period of rising cyclophosphamide Immunomodulation with Baseline phase anti-rhGAA IgG titers and rituximab bortezomib-based regimen Week 1 → week 38 Week 38 → week 86 Week 86 → week 98 of Prior to ERT of ERT of ERT ERT Week 98 of ERT onward Anti-rhGAA IgG titers Seronegative 1:25,600 → 1:102,400 1:102,400 → 1:204,800 1:204,800 1:204,800 → 1:100 Cardiac status 278.6 135 138.0 → 228.6 228.6 → 360.3 (With 360.3 → 72.0 (With no (LVMI (g/m2))a significant LVOTO; peak evidence of LVOTO) velocity 5.3 m/s) b c c d Urinary Glc4 39.9 NA 130.4 (Week 64) 153.3 (Week 92) 83.4 (Week 159) (mmol/mol creatinine) Gross motor status Generalized Sat unsupported; Regression of motor Voluntary motor Reacquisition of previously hypotonia rolled over; raised milestones; myopathic activity limited to ocular lost motor milestones; arms against gravity; facies movements only regained ability to move bore weight on lower fingers and toes, bend knees extremities and nod head; recovery of muscles of facial expression and improvement in speech function Respiratory status No respiratory No respiratory support Full-time ventilator Increase in ventilation Decrease in ventilation support dependent (PC/PEEP of requirement (PC/PEEP of requirement (PC/PEEP of

18/6 cm H2O, breathing 15/6 cm H2O, breathing 15/5 cm H2O, breathing rate rate: 6/min) (week 55 rate increased to 30/min) 10/min); ability to come onward) off ventilator frequently for short periods (~8–10 min) (week 159) Feeding status Oral Oral Gastrostomy tube Gastrostomy tube Gastrostomy tube (week 55 onward) Baseline: before initiation of ERT. ERT, enzyme replacement therapy; GAA, glucosidase α; LVMI, left-ventricular mass index; LVOTO, left-ventricular outflow tract obstruction; NA, not available; PC/PEEP,

pressure control/positive end-expiratory pressure; urinary Glc4, urinary glucose tetrasaccharide. aUpper limit of normal LVMI: 65 g/m2 (>2 SD higher than upper limit of the age-appropriate normal mean). bControl value for ages 1 to 6 months: 20 mmol/mol of creatinine (95th percentile). cControl value for ages 1 to 3 years: 8 mmol/mol of creatinine. dControl range for ages >3 years: 3 mmol/mol of creatinine.

B-lymphocyte CD19 count dropped from 17.9% at the onset bortezomib, in combination with rituximab, methotrexate, and of rituximab and methotrexate therapy at week 87 to 0.1% at IVIG, antibody titers decreased from 1:409,600 to 1:6,400. As week 89, antibody titers did not drop to <1:102,400. The anti- in patient 1, the decline in antibody titers in this bortezomib- body titers continued to rise again and remained at peak levels treated patient over time is in stark contrast to the persis- of 1:409,600 from week 143 through week 153. tence of HSATs in CRIM-negative patients (Supplementary At age 3 years and 2 months (week 146 of ERT; ~13 months Figure S1 online). Likewise, the observed reduction in anti- after the last dose of rituximab and 9 months after the last body titers, from 1:409,600 to 1:6,400, was closely associated dose of methotrexate), he developed an upper respiratory tract in time with clinical improvement, including cardiac, motor, infection, which led to hospitalization and intubation. During respiratory, and feeding status (Tables 2–4). He has had no his 6-week hospitalization, he was unable to be weaned from hospitalizations or illnesses since the start of the bortezomib- the ventilator; he required ventilator support at discharge based immunomodulatory regimen and has tolerated this (Tables 2–4). His cardiac, motor, and feeding status deterio- regimen well, with no apparent side effects. The patient con- rated while he was hospitalized, which was concurrent with tinues to receive alglucosidase alfa infusions at 20 mg/kg every increased urinary Glc4 levels (Tables 2–4). other week.

Immunomodulation using bortezomib-based regimen and Patient 3. Patient 3 is a 6-year and 10-month-old Caucasian subsequent clinical improvement. A trial of bortezomib was male diagnosed with atypical IPD at age 3.5 years based on commenced based on the following evidence: results of case reduced GAA activity in muscle as well as extensive cellular 1, worsening clinical status, lack of reduction of antibody vacuolization with glycogen accumulation. Patient 3 demo- titers below 1:102,400 with the rituximab and methotrexate graphics, CRIM status, and mutation data are shown in Table 1. regimen, and the known fact of poor outcome in patients with Symptoms began in the newborn period and included the find- HSATs. Details of the bortezomib-based immunomodulatory ings shown in Tables 2–4. Echocardiogram and electrocar- strategy are described in Figure 1b. Following four cycles of diogram findings remained within normal limits, confirming

Genetics in medicine | Volume 15 | Number 2 | February 2013 127 ORIGINAL RESEARCH ARTICLE BANUGARIA et al | Bortezomib in Pompe disease

Table 3 Clinical and laboratory parameters over time for the three patients treated with bortezomib-based regimen (patient 2) Patient 2 Immunomodulation attempt with Initial improvement Period of rising cyclophosphamide Immunomodulation with Baseline phase anti-rhGAA IgG titers and rituximab bortezomib-based regimen Week 1 → week 38 Week 38 → week 86 Week 86 → week 98 of Prior to ERT of ERT of ERT ERT Week 98 of ERT onward Anti-rhGAA IgG titers Seronegative 1:12,800 1:102,400 → 1:819,200 1:819,200 → 1:102,400 1:204,800 → 1:6,400 → 1:204,800 Cardiac status 200.9 95 95.0 → 100.0 100 → 60.3 → 139.0 139 → 82.5 (LVMI (g/m2))a b b b b b c Urinary Glc4 NA NA 16.8 → 23.2 23.2 → NA → 95.3 95.3 → 73.9 (mmol/mol creatinine) Gross motor status Generalized Clinical improvement Plateau of gross Regression of gross Stands with support; hypotonia; from baseline developmental progress motor milestones; remained increased extremity strength; head lag; facial wheelchair-bound; able to able to hold arms above body; myopathy; sit with support; unable to pushes against resistance bilateral stand unassisted ptosis; tongue protrusion Respiratory status No respiratory No respiratory support No respiratory support Required ventilator support Remains ventilator

support (PC/PEEP of 18/8 cm H2O, dependent; tolerates breathing rate 18/min) trach-mask trials off the (week 146) ventilator for 10–15 min twice a day; decrease in ventilation requirement

(PC/PEEP of 10/5 cm H2O, breathing rate 14/min) (week 221) Feeding status Oral Oral Oral Nasogastric tube Oral (week 146 onward) Baseline: before initiation of ERT. ERT, enzyme replacement therapy; GAA, glucosidase α; LVMI, left-ventricular mass index; NA, not available; PC/PEEP, pressure control/positive end-expiratory pressure;

urinary Glc4, urinary glucose tetrasaccharide. aUpper limit of normal LVMI: 65 g/m2 (>2 SD higher than upper limit of the age-appropriate normal mean). bControl value for ages 1 to 3 years: 8 mmol/mol of creatinine. cControl range for ages >3 years: 3 mmol/mol of creatinine. suspected atypical IPD. ERT with alglucosidase alfa was com- round of bortezomib (four doses) in combination with ritux- menced at age 3.6 years at 20 mg/kg every other week. At this imab, methotrexate, and IVIG, the antibody titer dropped from time, the patient required gastrostomy-tube feeds. Urine Glc4 1:204,800 to 1:51,200 (Figure 1c). Subsequent administration level at baseline was 67 mmol/mol of creatinine (control value of the bortezomib-based regimen resulted in further decline for age >3 years: 3 mmol/mol of creatinine). The patient showed in antibody titers to 1:3,200 at week 169 of ERT (Figure 1c). a good clinical response to ERT within the first 6–8 months Along with the drop in antibody titer, the patient experienced (26–34 weeks of ERT) as shown in Tables 2–4. significant clinical improvement. There was a substantial reduction in ventilator requirements, as well as an overall increase in Rising antibody titers. From a seronegative status at baseline, strength and energy, and significantly improved motor func- antibody titers rose to 1:204,800 at week 64 of ERT, where they tion (Tables 2–4). Improvements in speech and swallowing remained through week 90 (Figure 1c). Along with the rise in were also noted. Urine Glc4 levels continued to drop from a antibody titers, this patient started to decline clinically, includ- preimmunomodulatory value of 149.2 (week 88) to 66.9 mmol/ ing worsening of motor and respiratory status (Tables 2–4). mol creatinine at week 169 (Figure 1c). The treatment regimen

This was simultaneous with increased urinary4 Glc levels was well tolerated, with no evident side effects. The patient con- (Figure 1c, Tables 2–4). tinues to receive alglucosidase alfa infusions at 20 mg/kg every other week. Immunomodulation using bortezomib-based regimen and subsequent clinical improvement. Given the rise in antibody Discussion titers and concurrent clinical decline, as well as our acquired At present, there are over 75 therapeutic proteins approved by experience with the two cases mentioned above, a bortezomib- the Food and Drug Administration for life-threatening and based immunomodulatory regimen was initiated for patient 3 debilitating chronic diseases, with many more under develop- at week 88 of ERT. Details of the immunomodulatory strategy ment. A principal problem precluding the full clinical benefit using bortezomib are described in Figure 1c. Following the first derived from the use of these agents pertains to the elicitation

128 Volume 15 | Number 2 | February 2013 | Genetics in medicine Bortezomib in Pompe disease | BANUGARIA et al ORIGINAL RESEARCH ARTICLE

Table 4 Clinical and laboratory parameters over time for the three patients treated with bortezomib-based regimen (patient 3) Patient 3 Initial improvement Period of rising Immunomodulation with Baseline phase anti-rhGAA IgG titers bortezomib-based regimen Week 1 → week 34 Week 34 → week 88 Prior to ERT of ERT of ERT Week 88 of ERT onward Anti-rhGAA IgG titers Seronegative 1:12,800 → 1:51,200 1:51,200 → 1:204,800 1:204,800 → 1:3,200 Cardiac status WNL WNL WNL WNL (LVMI (g/m2))a b b b b Urinary Glc4 67 NA 111 → 149.2 149.2 → 66.9 (mmol/mol creatinine) Gross motor status Delayed gross Acquired new motor Less energy and less exercise Improved motor control, including hold- motor milestones, milestones; ability to ride tolerance than previously ing himself upright while in a stander; with inability to a stationary hand-pedal noted; able to stay in a stander able to stay in stander for up to 3 h ambulate or sit exercise machine up to for 25 min unsupported 15 min Respiratory status Required Improvement in ventilator Increased requirement for Decrease in ventilatory requirements continuous status (ability to spend ventilation; remained on with ability to come off of the ventilator ventilatory 1 h daily off ventilator) synchronized intermittent for short periods of time; remains on support mandatory ventilation without synchronized intermittent mandatory a backup breath rate for ventilation without a backup breath 1 hour; increased need for rate for the whole day; decreased tracheostomy suction requirement for tracheostomy suction Feeding status Gastrostomy tube Gastrostomy tube Gastrostomy tube Oral with improvement in swallow function plus supplementary gastrostomy tube feeds Baseline: before initiation of ERT.

ERT, enzyme replacement therapy; GAA, glucosidase α; LVMI, left-ventricular mass index; NA, not available; urinary Glc4, urinary glucose tetrasaccharide; WNL, within normal limits. aUpper limit of normal LVMI: 65 g/m2 (>2 SD higher than upper limit of the age-appropriate normal mean). bControl range for ages >3 years: 3 mmol/mol of creatinine. of immune responses, particularly when a therapeutic protein These case studies demonstrate both the safety and efficacy is immunologically perceived to be a foreign antigen. This rep- of bortezomib as an immunomodulatory agent in the setting resents a common challenge when considering treatment for of a well-established immune response to a therapeutic pro- patients with Pompe disease, mucopolysaccharidosis types I, tein. These are the first-known reported cases where successful II, and VI, Fabry disease, and hemophilias A and B.5,24,31 induction of a prolonged decline in HSATs in a disease with There are various factors that could explain an immune therapeutic protein has been achieved by the use of a protea- response to a therapeutic protein, but in the patients with some inhibitor (bortezomib). These cases demonstrate a direct Pompe mounting such responses to rhGAA, the major factor is relationship between the antibody response (titers and dura- the genetic status of the patient, in which severe mutations in the tion) and clinical response. In these three cases, treatment GAA gene either fail to produce any enzyme or produce a defec- initiation with bortezomib was rapidly followed by sustained tive enzyme that fails to tolerize the patients’ immune systems. reductions in antibodies and clinical benefit. The rapid reduc- Thus, to such an untolerized immune system, the full-length tion in antibody titers occurred within a few weeks of start- human GAA appears as a foreign protein, to which an immune ing the bortezomib-based regimen, with titers dropping from response is mounted.24 Other factors of importance in elicitation 1:204,800 to 1:100; 1:409,600 to 1:6,400; and 1:204,600 to 1:3,200 of immune responses to therapeutic proteins include the follow- in patients 1, 2, and 3, respectively. This represents a 2,048-fold ing: structural properties of a protein (e.g., sequence variation (patient 1) and 64-fold (patients 2 and 3) decline in titers as and glycosylation); impurities and contaminants; frequency and compared with titers at the time bortezomib was initiated. Of duration of treatment; genetic background of patients; route of note, the marked and sustained decrease in antibody titers was administration; and other host and environmental factors.32 associated with significant, durable improvement across all In patients who develop HSATs against ERT, not only can the clinical outcome measures with continued improvement at the prognosis be poor but sometimes it can be fatal.3,4,31 In addition time of publication. Clearly, the benefits have been more robust to the human toll, the use of therapeutic proteins in patients in cardiac parameters as compared with the skeletal muscle who are not responding to the treatment because of interfer- response in patients 1 and 2. This is probably due to irreversible ing antibodies can have a large economic impact. Investigation skeletal muscle damage that has been noted in other infantile of novel immunomodulatory strategies to preclude or reverse survivors, despite long-term treatment with ERT.12,33 immune responses—and to induce immune tolerance in this Urinary Glc4 levels correlate with overall glycogen burden setting—is not only critical in terms of therapeutic effect but and are useful for monitoring response to ERT.3,4,30. For all also for optimal use of health-care resources. three patients, the increase in antibody titers correlated with an

Genetics in medicine | Volume 15 | Number 2 | February 2013 129 ORIGINAL RESEARCH ARTICLE BANUGARIA et al | Bortezomib in Pompe disease

increase in urinary Glc4 levels and clinical decline. Conversely, and Becker muscular dystrophies as well as in rat models of the decrease in antibody titers associated with immunomodu- denervation-induced muscular atrophy.37,38 lation using the bortezomib-based regimen led to a reduction Recent reports have demonstrated the successful use of

(albeit of variable magnitude) in urinary Glc4 levels and clinical immunomodulatory regimens based on rituximab, metho- improvement (Tables 2–4). trexate, and either with or without IVIG, in the ERT-naive or As with any immunosuppressive therapy, vaccination response early-ERT settings.17,18 At present, we do not have enough evi- might be diminished with the described bortezomib-based dence for the use of the bortezomib-based immunomodulatory regimen. Live vaccines should be avoided while on treatment regimen described here in the ERT-naive or early-ERT settings. and immediately following treatment. Bortezomib is associated Conservatively, it appears that this bortezomib-based immuno- with peripheral neuropathy, neutropenia, thrombocytopenia, modulatory regimen should be used in patients with (i) anti- and gastrointestinal and cardiac side effects. However, no side body titers of ≥51,200 on two or more occasions at or beyond effects were evident in any of these three patients treated with 6 months of ERT, as they seem to be at highest risk of clinical the bortezomib-based regimen. Total levels of different isotypes decline following the development of HSATs4 and/or (ii) an of immunoglobulin (total IgG, IgA, IgE, IgM, and IgD) stayed increasing trend or persistence of antibody titers (regardless of within normal ranges while on this regimen, except for two titer value) with associated clinical decline. instances in patient 2 and one instance in patient 1, wherein As used in these three patients, a combination of drugs that IgA levels were above normal limits (data not shown). Given target B cells, T cells, and plasma cells is needed to achieve the dramatic suppression of immunity, the patients were sup- long-term remission of antibodies and potential induction of plemented with monthly IVIG for infection prophylaxis and to immune tolerance, as depicted in Supplementary Figure S2 potentially contribute to immunomodulation.34 online. In these cases, the combination of rituximab, metho- Although the mechanism of action of bortezomib in reduc- trexate, and IVIG in addition to bortezomib appeared to be both ing anti-rhGAA IgG antibodies is not precisely known in these safe and efficacious, with each drug acting on a different part of cases, we speculate that it is due to proteasome inhibition in the pathway leading to antibody production (Supplementary both short-lived and long-lived plasma cells. Bortezomib binds Figure S2 online). to the catalytic site of the 26S proteasome with high affinity The immunomodulatory approach described herein repre- and specificity. There are various explanations for the actions sents a promising new strategy for patients with entrenched of bortezomib and other proteasome inhibitors on plasma cells: immune responses to protein replacement therapies. It may (i) plasma cells are exceptionally sensitive to proteasome inhibi- also prove successful in addressing autoimmune disorders that tors due to their excessive protein (immunoglobulin) synthesis. fail to respond to agents targeting only T and B cells, such as Blocking the proteasome rapidly induces excessive endoplas- rituximab and methotrexate. Thus, this immunomodulatory mic reticulum stress and activation of the terminal unfolded approach should be considered for further study in a variety protein response with induction of the proapoptotic protein of settings where patient immune response is implicated in the and activation of caspases, leading to apoptosis;26 (ii) protea- disease pathogenesis itself or in reduced effectiveness of the some inhibition interferes with degradation of IκB (nuclear fac- respective protein therapy. tor of κ light polypeptide gene enhancer in B-cell inhibitors) proteins, thereby blocking activation of the transcription fac- Supplementary material tor NF-κB (nuclear factor κ-light-chain-enhancer of activated Supplementary material is linked to the online version of the paper B-cells), which could contribute to bortezomib-induced plasma at http://www.nature.com/gim cell death;26 and (iii) bortezomib alters the microenvironment (e.g., interleukin-5 and interleukin-6, tumor necrosis factor-α) ACKNOWLEDGMENTS required for plasma cell survival and could play a role in its The authors thank the patients who participated in this clinical elimination.26 In addition to its primary action on plasma cells, study and their families. They also thank the study coordinators, bortezomib has also been found to act on both T and B cells, nurses, and technical staff at the various treatment centers. The thereby further modulating the humoral antibody response.35 opinions and information in this article are those of the authors, As a proteasome inhibitor, bortezomib may also directly act and do not necessarily represent the views and/or policies of the on muscle in the context of IPD. The ubiquitin-proteasome US Food and Drug Administration. system is believed to degrade the major contractile skeletal muscle proteins and to play a critical role in muscle wasting. Disclosure Muscle wasting is a prominent feature of IPD and is similar to P.S.K. reports receiving research and grant support from Genzyme. other muscle wasting disease states, such as cancer cachexia, P.S.K. also receives honoraria and consulting fees from Genzyme sepsis, diabetes, and metabolic acidosis, in which expression of and is a member of the Pompe disease and the Gaucher Disease the ubiquitin-proteasome proteolytic pathway is increased in Registry Advisory Boards. Duke University and the inventors of the skeletal muscle.8,36 Thus, blockade of the proteasome in muscle method of treatment and precursors of the cell lines used to gen- has the potential to mitigate skeletal muscle wasting and dam- erate the enzyme (rhGAA) used commercially have received royal- age, which has been demonstrated preclinically in Duchenne ties pursuant to the University’s policy on inventions, patents, and

130 Volume 15 | Number 2 | February 2013 | Genetics in medicine Bortezomib in Pompe disease | BANUGARIA et al ORIGINAL RESEARCH ARTICLE technology transfer. This potential conflict for Duke University has 18. Mendelsohn NJ, Messinger YH, Rosenberg AS, Kishnani PS. Elimination been resolved through monetization. S.G.B., S.N.P., and P.S.K. are of antibodies to recombinant enzyme in Pompe’s disease. N Engl J Med 2009;360:194–195. listed as inventors on a Duke University pending patent applica- 19. Hunley TE, Corzo D, Dudek M, et al. Nephrotic syndrome complicating tion for the work described in this article. To date, neither Duke alpha-glucosidase replacement therapy for Pompe disease. Pediatrics University nor the inventors have received any money from rights 2004;114:e532–e535. associated with this pending patent. D.V. receives clinical research 20. Banugaria SG, Patel TT, Mackey J, et al. Persistence of high sustained antibodies to enzyme replacement therapy despite extensive immunomodulatory therapy support from Genzyme and serves as a board member for the in an infant with Pompe disease: need for agents to target antibody-secreting MPS I Registry in which honoraria are provided for attendees plasma cells. Mol Genet Metab 2012;105:677–680. at meetings by the host company, Genzyme. The other authors 21. Biss TT, Velangi MR, Hanley JP. Failure of rituximab to induce immune tolerance in a boy with severe haemophilia A and an alloimmune factor VIII declare no conflict of interest. antibody: a case report and review of the literature. Haemophilia 2006;12: 280–284. REFERENCES 22. Hruby MA, Schulman I. Failure of combined factor VIII and cyclophosphamide to 1. Porter S. Human immune response to recombinant human proteins. J Pharm Sci suppress antibody to factor VIII in hemophilia. Blood 1973;42:919–923. 2001;90:1–11. 23. Collins PW, Mathias M, Hanley J, et al. Rituximab and immune tolerance in 2. Amalfitano A, Bengur AR, Morse RP, et al. Recombinant human acid alpha- severe hemophilia A: a consecutive national cohort. J Thromb Haemost 2009;7: glucosidase enzyme therapy for infantile glycogen storage disease type II: 787–794. results of a phase I/II clinical trial. Genet Med 2001;3:132–138. 24. Wang J, Lozier J, Johnson G, et al. Neutralizing antibodies to therapeutic 3. Kishnani PS, Goldenberg PC, DeArmey SL, et al. Cross-reactive immunologic enzymes: considerations for testing, prevention and treatment. Nat Biotechnol material status affects treatment outcomes in Pompe disease infants. Mol Genet 2008;26:901–908. Metab 2010;99:26–33. 25. Richardson PG, Barlogie B, Berenson J, et al. A phase 2 study of bortezomib in 4. Banugaria SG, Prater SN, Ng YK, et al. The impact of antibodies on clinical relapsed, refractory myeloma. N Engl J Med 2003;348:2609–2617. outcomes in diseases treated with therapeutic protein: lessons learned from 26. Neubert K, Meister S, Moser K, et al. The proteasome inhibitor bortezomib infantile Pompe disease. Genet Med 2011;13:729–736. depletes plasma cells and protects mice with lupus-like disease from nephritis. 5. Hollak CE, Linthorst GE. Immune response to enzyme replacement Nat Med 2008;14:748–755. therapy in Fabry disease: impact on clinical outcome? Mol Genet Metab 27. Kishnani PS, Nicolino M, Voit T, et al. Chinese hamster ovary cell-derived 2009;96:1–3. recombinant human acid alpha-glucosidase in infantile-onset Pompe disease. 6. Kishnani PS, Hwu WL, Mandel H, Nicolino M, Yong F, Corzo D. A retrospective, J Pediatr 2006;149:89–97. multinational, multicenter study on the natural history of infantile-onset Pompe 28. Klinge L, Straub V, Neudorf U, et al. Safety and efficacy of recombinant disease. J Pediatr 2006;148:671–676. acid alpha-glucosidase (rhGAA) in patients with classical infantile Pompe 7. van den Hout HM, Hop W, van Diggelen OP, et al. The natural course of infantile disease: results of a phase II clinical trial. Neuromuscul Disord 2005;15: Pompe’s disease: 20 original cases compared with 133 cases from the literature. 24–31. Pediatrics 2003;112:332–340. 29. Velcade. [Package Insert]. Millenium Pharmaceuticals: Cambridge, MA, 8. Hirschhorn R, Reuser AJJ. Glycogen storage disease type II: acid a-glucosidase 2010. (acid maltase) deficiency. In: Valle D, Scriver CR, eds. Scriver’s Ommbid The 30. Young SP, Piraud M, Goldstein JL, et al. Assessing disease severity in Pompe Online Metabolic & Molecular Bases of Inherited Disease. McGraw-Hill: New disease: the roles of a urinary glucose tetrasaccharide biomarker and imaging York, 2009. http://genetics.accessmedicine.com/. techniques. Am J Med Genet C Semin Med Genet 2012;160:50–58. 9. Kishnani PS, Steiner RD, Bali D, et al. Pompe disease diagnosis and management 31. Ragni MV, Ojeifo O, Feng J, et al. Risk factors for inhibitor formation guideline. Genet Med 2006;8:267–288. in haemophilia: a prevalent case-control study. Haemophilia 2009;15:1074– 10. Kishnani PS, Corzo D, Nicolino M, et al. Recombinant human acid (alpha)- 1082. glucosidase: major clinical benefits in infantile-onset Pompe disease. Neurology 32. Schellekens H. Factors influencing the immunogenicity of therapeutic proteins. 2007;68:99–109. Nephrol Dial Transplant. 2005;20(suppl 6):vi3–9. 11. Nicolino M, Byrne B, Wraith JE, et al. Clinical outcomes after long-term 33. Prater SN, Banugaria SG, Dearmey SM, et al. The emerging phenotype treatment with alglucosidase alfa in infants and children with advanced Pompe of long-term survivors with infantile Pompe disease. Genet Med 2012;14: disease. Genet Med 2009;11:210–219. 800–810. 12. Kishnani PS, Corzo D, Leslie ND, et al. Early treatment with alglucosidase 34. Hall PD. Immunomodulation with intravenous immunoglobulin. alpha prolongs long-term survival of infants with Pompe disease. Pediatr Res Pharmacotherapy 1993;13:564–573. 2009;66:329–335. 35. Lemy A, Toungouz M, Abramowicz D. Bortezomib: a new player in 13. de Vries JM, van der Beek NA, Kroos MA, et al. High antibody titer in an adult pre- and post-transplant desensitization? Nephrol Dial Transplant 2010;25: with Pompe disease affects treatment with alglucosidase alfa. Mol Genet Metab 3480–3489. 2010;101:338–345. 36. Tisdale MJ. Is there a common mechanism linking muscle wasting in various 14. Patel TT, Banugaria SG, Case LE, Wenninger S, Schoser B, Kishnani PS. The disease types? Curr Opin Support Palliat Care 2007;1:287–292. impact of antibodies in late-onset Pompe disease: A case series and literature 37. Gazzerro E, Assereto S, Bonetto A, et al. Therapeutic potential of proteasome review. Mol Genet Metab 2012;106:301–309. inhibition in Duchenne and Becker muscular dystrophies. Am J Pathol 15. Myozyme. [Package Insert]. Genzyme Corporation: Cambridge, MA, 2006. 2010;176:1863–1877. 16. Banati M, Hosszu Z, Trauninger A, Szereday L, Illes Z. Enzyme replacement 38. Beehler BC, Sleph PG, Benmassaoud L, Grover GJ. Reduction of skeletal muscle therapy induces T-cell responses in late-onset Pompe disease. Muscle Nerve atrophy by a proteasome inhibitor in a rat model of denervation. Exp Biol Med 2011;44:720–726. (Maywood) 2006;231:335–341. 17. Messinger YH, Mendelsohn NJ, Rhead W, et al. Successful immune tolerance 39. Vogel M, Staller W, Bühlmeyer K. Left ventricular myocardial mass determined induction to enzyme replacement therapy in CRIM-negative infantile Pompe by cross-sectional echocardiography in normal newborns, infants, and children. disease. Genet Med 2012;14:135–142. Pediatr Cardiol 1991;12:143–149.

Genetics in medicine | Volume 15 | Number 2 | February 2013 131